One of the more remarkable feats of neural circuits is the extraction of information from light. The vertebrate retina receives, processes and transforms visual information. Light absorbed in photoreceptors is transformed into membrane polarizations that evoke light modulated release of the neurotransmitter glutamate onto horizontal and bipolar cells (HCs and BCs). HCs feed back onto photoreceptor synapses, regulating transmitter release according to conditions of ambient illumination. BCs transfer light signals forward to amacrine cells (ACs) and ganglion cells (GCs). ACs are mainly regulatory retinal interneurons with inhibitory and modulatory roles. GCs transform light signals into trains of action potentials that propagate through the optic nerve to brain visual centers. As images pass through retinal circuitry, they are decomposed into component parts, so that at the retinal output specialized sets of GCs report different features of the image. Some signal highlights, others shadows, movement, direction or color. This is the image processing done by the retina. It is achieved by sets of specialized neural circuits. Such circuits are composed of patterns of connections among neurons, and specialized interactions between neurotransmitters and their receptors. This research program studies the relationships between receptor expression on retinal neurons, the neural circuitry of the retina, and retinal information processing tasks. Zebrafish is a visual system model that provides behavioral, molecular, anatomical and physiological measures. Through studies of this model, genetic and other perturbations of the visual process may be better understood. We examine zebrafish retinal function using acutely dissociated retinal neurons, in vitro eyecup preparations, and zebrafish retinal slices. These preparations provide infomation about structural neural pathways, the responses of neural types to stimulaton by neurotransmitters, and by light. Through integrating and synthesizing such information, it is possible to infer the physiological transformations that the retina imposes on the image. Cell structure, a key element in discerning retinal circuits, is beautifully delineated in the zebrafish model. Neurons in retinal slices can be stained in isolation shotgun fashion by spraying DiI coated microcarriers onto the cut surface with a gene gun. This is the Diolisitic technique. Diffusion of the lipophilic tracer along membranes reveals the shapes of individual neurons. In a study of 300 stained zebrafish retinal neurons different subtypes of HCs, BCs and ACs were observed. Based on cell body shape, and the presence or absence of an axon, 3 HC morphologies were identified corresponding to stellate and elongate types earlier seen in cell culture. Based on branching patterns of axons and dendrites within the retinal inner plexiform layer 17 BC morphologies and 7 AC morphologies were discerned. Cells with ON-type (branching in IPL sublamina b), OFF-type (branching in sublamina a) morphological signatures were about equally prevalent. Mixed a/b branching patterns were also observed. In previous studies using Lucifer-dye filled patch electrodes 13 morphological types of bipolar cell were identified. There is substantial agreement on morphological types between the two studies. Whole-cell patch recording and puff pipette techniques have identified glutamate receptor mechanisms on the BC dendrites of many of these types showing a distribution of 3 basic glutamate receptor types: AMPA/kainate (OFF cells), mGluR6 (ON cells) and glutamate-gated chloride currents (Iglu, ON cells). Diolisitic studies, however suggest that in addition to 3 distinct strata within sublamina a of the IPL, and 2 within sublamina b, there is also a third stratum of sublamina b, intermixed with the retinal ganglion cell layer. The 3 distinct glutamate receptors expressed on bipolar cells can be isolated in the light-evoked field potential of the zebrafish eye. These field potentials are called electroretinograms or ERG's. CNQX, an antagonist of AMPA/kainate glutamate receptors blocks the light responses of OFF type bipolar cells and the d-wave of the ERG. The d-wave is a transient vitreal postive potential seen at light offset. The b-wave can be blocked by TBOA, an antagonist of glutamate transporters. Glutamate transporters generate the responses of many ON-type bipolar cells, as these transporters combine to form chloride channels when activated by photoreceptor glutamate. The combination of CNQX and TBOA nearly abolishes the ERG, suggesting that these two components are predominant. However if the metabotropic (mgluR6) glutamate-receptor-blocker DL-AP4 is added to this mix, a sustained, light evoked, vitreal negative response appears. This is evidentally the photoreceptor component, as all known glutamate receptor mechanisms post synaptic to photoreceptors are blocked by the combination of these three antagonists. Evidentally the metabotropic ON-bipolar ERG component and photorecptor componenet approximately cancel, so that with only these two components functional, little light response is seen. ON and OFF mechanisms, as seen in the ERG b- and d-waves, appear to interact. Suppression of the d-wave by CNQX increases the magnitude of b-wave responses, and this is true of both metabotropic and transporter driven components. Similarly, suppression of either metabotropic or glutamate transporter driven components of the b-wave increase d-wave amplitudes. The d-wave enhancement produced by TBOA transporter blockade can be blocked by CNQX. The d-wave enhancement produced by metabotropic blockade with DL-AP4 survives CNQX treatment. The mutual antagonism between amplitudes of b- and d-waves suggest circuitry interactions between ON and OFF bipolar cells. ERG responses from zebrafish retina appear rich in information about the physiological processes involved in image processing within the retinal neural circuit.